Although epidemiologic evidence indicates that most human cancers are caused by unknown exposures to carcinogens, epidemiologists still rely on self-reported information to characterize exposures, much as they did a century ago. To provide high-quality data regarding carcinogen exposures from diverse exogenous and endogenous sources, studies of cancer etiology should use untargeted omics of chemicals in blood. Unfortunately, many carcinogens cannot be measured in blood because they are reactive electrophiles with short life spans in vivo. An avenue for assessing these meaningful exposures focuses on adducts from reactions between reactive electrophiles and blood proteins. Adducts in human serum albumin (HSA) are particularly appealing because HSA is the most abundant protein in serum and Cys34 is the dominant scavenger of reactive electrophiles in serum. Other nucleophilic hotspots of HSA include Lys199 and His146. Although levels of some HSA adducts have been correlated with exposures to targeted chemicals, a plethora of uncharacterized adducts represent unknown exposures that undoubtedly contribute to cancer incidence. Thus the `HSA adductome', representing the totality of such adducts, is of great potential importance to cancer epidemiology. Since all reactive electrophiles are possible carcinogens, measurement of the HSA adductome is arguably more relevant to the discovery of initiators of human cancers than the proteome or the metabolome, both of which are being applied to investigate cancer causation. With previous NIH funding, we used triple-quadrupole mass spectrometry (MS) to perform untargeted analyses of Cys34 adducts with samples of human serum. Although that project provided proof-of-concept, coverage was limited to Cys34 modifications and triple-quadrupole MS did not provide accurate masses for annotation. We recently showed that liquid chromatography-high resolution mass spectrometry was more appropriate for adductomics and developed a robust `adductomics pipeline' for characterizing Cys34 adducts. The goals of the proposed project are to expand coverage of HSA adductomics by simultaneously measuring adducts of Cys34, Lys199 and His146. After improving the technology, we will validate methods with specimens of plasma from healthy adults and will pilot-test the technology for cancer epidemiology with archived serum from incident cases of non-Hodgkin's lymphoma (NHL) and matched controls from the EPIC- Italy prospective cohort study. Successful innovations of this technology can transform approaches to discover the origins of human cancers.